New The SREYUN website is live: five in-house products and reference pricing Learn more →

LLM API cost allocation and budget control

AI infrastructure For: Engineering and finance teams using LLMs on several clouds Products: TokenTally

Bring the API keys spread over several Alibaba Cloud Model Studio and Tencent Cloud TokenHub accounts together, allocate costs to departments, set budgets and hear about cost anomalies the same day.

The challenge

  • Support, marketing, engineering and data teams each request API keys, spread over several cloud accounts and workspaces.
  • Clouds bill by product, so finance cannot allocate costs to departments and projects.
  • A leaked or misused key is often only noticed when the bill arrives at month end.

The solution

Deploy TokenTally with read-only cloud keys:

  • One view: metadata syncs every 5 minutes and bills every 60 minutes; key-level costs from both clouds are calculated the same way and match the bills.
  • Cost allocation: give keys a cost center and an owner, pivot cost center × model and export.
  • Budgets and alerts: monthly total, department and key budgets; cost anomalies are judged against the average list price of the previous 7 days and reported the same day.
  • Governance: automatically finds keys that are deleted but still billed, over-privileged or idle.

The value

  • What every key, model and department spends, on one page.
  • Daily, weekly and monthly reports are sent automatically with a sign-in-free read-only interactive report, so owners can check any time.
  • Read-only permissions and local storage: nothing in your clouds is changed.